The nonobese diabetic (NOD) mouse represents an excellent model for analyzing pathogenic mechanisms underlying autoimmune destruction of pancreatic beta cells. Our objectives are to elucidate the immunopathogenic basis underlying the failure of NOD T cells to develop tolerance to beta cell antigens, and to use gene therapy to restore more normal immunoregulation and thus prevent autoimmune diabetes. We have shown that pathogenesis initiates at the hematopoietic stem cell level since NOS bone marrow adoptively transfers disease into otherwise diabetes-resistant F1 radiation chimeras. We have established that although diabetes in NOD mice is under polygenic control, the MHC class II region of the unique H-2 haplotype encodes a major component of susceptibility. We have identified a number of defects in signaling between NOD antigen presenting cells (APC) and T cells. Some of these defects express as an aberrant development of marrow-derived macrophages (MO). We found that APC transgenically expressing an H-2-Ea class II transgene block the development of diabetogenic T cells from NOD marrow. This renewal proposal has three objectives. The first is to develop a gene therapy protocol t reverse the diabetogenic potential of NOD marrow. We propose to use retroviral vectors to transduce a protective H-2-Ea cDNA ligated to a genomic promoter into NOD hematopoietic stem cells. We shall test whether APC derived from such retrovirally-transduced stem cells express sufficient I-E molecules to block diabetes. Our second specific aim will employ competitive bone marrow chimeras and fetal thymic organ culture to test whether aberrant MO development also underlies failure of NOD T cells to become tolerant to beta cells. The third specific aim is to use intrathymic injection of candidate Beta cell autoantigens into adolescent NOD mice as a means for identifying the primary autoantigens to which tolerance must be induced to circumvent diabetogenesis. Successful completion of these studies will not only provide a gene therapy model for autoimmune insulin dependent diabetes in humans, but may also elucidate immunophenotype that identify prediabetic humans who may be the logical recipients of this therapy.